The mass analysis of large or macro molecules was a difficult task prior to the advent of then Electrospray [ES] Ionization technique which is described in a number of U.S. Pat. Nos. (Labowsky et al., 4,531,056; Yamashita et al., 4,542,293; Henion et al. 4,861,988; and Smith et al. 4,842,701 and 4,887,706) and in several recent review articles [Fenn et al., Science 246, 64 (1989); Fenn et al., Mass Spectrometry Reviews 6, 37 (1990); Smith et al., Analytical Chemistry 2, 882 (1990)]. Because of extensive multiple charging ES ions of large molecules almost always have mass/charge (m/z) ratios of less than about 2500 so they can be weighed with relatively simple and inexpensive conventional analyzers.
The multiple charging characteristic of the ES and other ion sources was originally viewed by workers in the field as a detriment. Indeed, mass spectrometrists were accustomed to analyzing spectrum in which each molecule was singly charged. The multiply charged spectrum were looked at as interesting curiosities until Mann. et.al [Interpreting Mass Spectra of Multiply Charged Ions, Anal. Chem. 1989, 61, 1702-1708} and Fenn et.al. (U.S. Pat. No. 5,130,538] revealed a algorithm which transformed the sequence of peaks for a multiply charged ion in the measured spectrum into a "calculated" (also referred to as "deconvoluted") spectrum in which all peaks represented singly charged parent ions. The Mann and Fenn algorithm is based on the fact that there are three unknowns associated with the ions of a particular peak in an measured spectrum: the molecular weight Mr of the parent species, the number i of charges on the ion, and the mass m.sub.a of each adduct charge. Therefore, mass/charge (m/z) values for the ions of any three peaks of the same parent species would fix the values of each unknown. However, there is a relation between the peaks such that they form a sequence, referred to as a "coherent" sequence, in which the number of charges i varies by one from peak to peak. Consequently, the m/z values of any pair of peaks are sufficient to fix Mr for the parent species, provided that the masses of the adduct charges are the same for all ions of all the peaks in the sequence. Using this information, Mann-Fenn devised a summing procedure which deconvoluted the measured spectrum. Indeed, the Mann and Fenn "deconvolution" method allowed for the extremely accurate determinations of molecular weights of very large molecules and greatly expanded the field of mass spectrometry.
In spite of the effectiveness of the Mann-Fenn deconvolution method, as originally described, it suffered from several disadvantages. The most obvious of these disadvantages was the calculated spectrum was very noisy and contained artifact peaks which made the identification of secondary species difficult. The Mann-Fenn method also suffered from a "high mass" bias. In other words, the method tended to attribute a higher signal to larger molecular weight ions. Also, one must known or have a reasonable estimate of the adduct ion mass before implementing the method. In using the Mann-Fenn algorithm one must assume an adduct ion mass. The calculation is then performed in which only the mass of the macromolecule is the only independent variable. As a result, the Mann-Fenn algorithm produces a 2-Dimensional (2-D) spectrum of calculated signal versus mass. Indeed, in principle, the adduct ion mass should be known before hand. In practice, the adduct ion mass may not be known for a number of reasons. First, there may not be a single adduct ion mass, but several such ions which attach to the parent molecule. Second, one may simply guess wrong when assigning the adduct ion mass. If one were to guess that the adduct ion were 1 (for a proton) and the real adduct ion were a protonated water (ma=19), the result obtained by the Mann-Fenn algorithm would be in gross error. Finally, even if the there were only one adduct ion and the user was certain of the mass of that ion, the result obtained by Mann-Fenn could still be in error due to the lack of proper calibration of the mass spectrometer which was used to generate the original spectrum. For these reasons as was pointed out in Labowsky [U.S. Pat. No. 5,300,771] and Labowsky et.al. [Rapid Comm in Mass Spectro, Vol 7, PP 71-84 (1993)] it is best to treat the adduct ion mass as an unknown. The result of such an approach is a 3-Dimensional (3-D) surface of calculated signal versus macromass and adduct ion mass. It should be mentioned that a 2-D calculated spectrum is simply a cross-section of a 3-D calculated surface at a given value for the adduct ion mass.
Noise reduction in the calculated spectrum is important whether the calculation is performed in two dimensions or in three dimensions. If noise and artifact peaks could be reduced or eliminated, then it would be easier to identify the masses of all species that may be represented an measured spectrum. FIG. 1 shows a measured spectrum of Cytochrome C. FIG. 2 shows a 2-D (cross-section) calculation of this spectrum at an adduct ion mass of 1 using the Mann-Fenn algorithm with no noise reduction. FIG. 3 shows the 3-D calculated surface of this spectrum, again with no noise reduction. It is quite clear from an examination of these figures that the presence of secondary and tertiary species which may be present in the measured spectrum may be obscured due to noise and artifact peaks. Conversely, an examination of a noisy calculated spectrum may lead one to conclude that a certain species is present in a measured spectrum if an artifact peak appears at the mass corresponding to that species in the calculated spectrum. For these reasons it is important to develop techniques for reducing the noise in the calculated spectra whether they are 2-D or 3-D.
Attempts to reduce noise and artifact peaks in the basic 2-D algorithm were made by Zhou (U.S. Pat. No. 5,072,115). Using what may be described as a peak subtraction method, Zhou used an iterative method to produce a calculated spectrum from an measured spectrum. In the first step of this iteration, the Mann-Fenn algorithm is applied to the measured spectrum to find the mass of the dominate species represented in the measured spectrum. In the second iteration, the Mann-Fenn algorithm is then applied to the measured spectrum in which the peaks associated with the dominant species were, by some means, subtracted out. In so doing, Zhou was able to calculate the mass of the next species. In the third iteration the Mann-Fenn algorithm again applied to a spectrum in which the peaks of the dominant and the second species are subtracted out to find the third species and so on. The process of identifying species and subtracting their peaks out from the measured spectrum is then repeated until all species have been identified. A final calculated spectrum is presented which has less noise and fewer artifact peaks than that produced by a single iteration of the Mann-Fenn algorithm.
While the method of Zhou represents an improvement of the basic Mann-Fenn algorithm, it involves several iterations to obtain a final solution. Further, it has been applied to obtain only 2-D calculated spectra. An application of this method to obtain a 3-D surface would be tedious in light of the large number of calculations that an iterative method would require to generate a 3-D surface.